The D-Dimer to Albumin Ratio Could Predict Hospital Readmission Within One Year in Patients with Acute Exacerbation of Chronic Obstructive Pulmonary Disease

Introduction

Chronic obstructive pulmonary disease (COPD) is a prevalent and debilitating respiratory condition, characterized by persistent airflow restriction and frequent exacerbations.¹ COPD ranks as the third leading cause of mortality globally, resulting in over 3 million deaths annually.^2,3 In China, COPD affects approximately 13.6% of adults aged over 40 years, imposing a significant disease burden.⁴ Acute exacerbation of COPD (AECOPD), a condition in which respiratory symptoms increase beyond the level of daily variation and necessitate changes in treatment, is a major contributor to hospitalizations and increased healthcare costs.⁵ The incidence of acute exacerbation could hasten disease progression, diminish quality of life, and increase mortality risk.⁶ Approximately 20% of patients undergo hospital readmission for AECOCP within 30 days after admission.^7,8 And more than 50% of patients experience at least one readmission for acute exacerbation within 1 year after discharge, with frequent exacerbation correlating with progressive COPD severity and increased mortality risk.^9,10 Despite advancements in COPD management, the prediction and prevention of AECOPD hospital readmissions remain a challenge. Identifying indicators that can predict AECOPD exacerbation and subsequent hospital readmission is crucial for improving patient prognosis.

The D-dimer-to-albumin ratio (DAR) is a sensitive indicator of hypercoagulability, inflammation, and nutritional status, and has been shown to have prognostic value in COVID-19 and gastric cancer.^11,12 DAR may also be linked to disease severity and prognosis in patients with COPD. COPD is a chronic inflammatory disease characterized by a hypercoagulable state.¹³ And an interactive relationship exists between coagulation and inflammation.^14,15 D-dimer not only reflects the activation of the coagulation system, but also promotes the release of inflammatory factors and contributes to the inflammatory response.^16,17 In stable COPD, elevated D-dimer levels are associated with increased disease severity and mortality risk.^18,19 During acute exacerbations, high D-dimer levels are linked to an increased risk of hospitalized and one-year mortality in patients with AECOPD.²⁰ Low levels of albumin, an acute phase response protein, may signify persistent inflammation in AECOPD.²¹ Malnutrition, as indicated by albumin levels, is related to more frequent acute exacerbations in COPD patients.²² Additionally, low serum albumin levels are associated with prolonged hospitalization in patients with AECOPD.²³

Our study aimed to explore the association between DAR and hospital readmission within one year in patients with AECOPD. The findings would contribute to risk stratification of COPD patients, enabling the timely identification of high-risk patients and the implementation of targeted interventions to slow disease progression and improve prognosis.

Methods

Study Design, Setting and Participants

All patients with COPD during the period from January 2019 to October 2022 in Baise People’s Hospital, Guangxi, were included in this retrospective cohort study. Cases were included if the patients were older than 18 years, in hospitalization for more than 24 hours,²⁴ have a complete electronic patient documents, and had a confirmed diagnosis of AECOPD. AECOPD was defined as an acute worsening of respiratory symptoms that results in additional therapy.⁵ The exclusion criteria were (1) have died during hospitalization, (2) receiving glucocorticoid therapy before blood sampling, and (3) missing data on D-dimer, albumin, high sensitivity C-reactive protein (hsCRP), and blood urea nitrogen. The study was conducted according to the Declaration of Helsinki, and the procedures were approved by the Ethics Committee of the Baise People’s Hospital (approval number: LW2024060401). The need for written informed consent was waived by the Ethics Committee of the Baise People’s Hospital due to retrospective nature of the study.

Data Collection

Patient data were recorded from the electronic patient records and managed using the hospital information system. The collection was performed by two doctors and a nurse. Demographic and clinical values, including age, sex, drug used, smoking, body mass index (BMI), heart failure, type 2 diabetes mellitus, hypertension, asthma, pneumonia, hemoglobin, DAR, neutrophil-to-lymphocyte ratio (NLR), D-dimer, albumin, red blood cell, white blood cell, platelet, monocyte, eosinophils, red blood cell distribution width (RDW-CV), alanine transaminase, aspartate transaminase, glutamyltransferase, blood urea nitrogen, creatinine, prothrombin time, activated partial thromboplastin time, thrombin time, fibrinogen, hsCRP, the forced expiratory volume in one second/the forced vital capacity (FEV1/FVC), treat with glucocorticoid/antibiotic/mechanical ventilation, inpatient time, uric acid, PaO₂, PaCO₂, SaO₂, PH, BMI, comorbidity, times of acute exacerbation in previous one year, modified Medical Research Council (mMRC), COPD assessment test (CAT), COPD course, were recorded at admission. If a patient has two or more of the five comorbidities of heart failure, diabetes, hypertension, asthma and pneumonia, the individual was considered multimorbidity, otherwise the individual was non-multimorbidity. The mMRC scale was used to assess the dyspnea of patients, with higher grade indicating more severe dyspnea of patients.^25,26 The CAT was to supplement information obtained from lung function measurement and assessment of exacerbation risk.²⁷ The CAT consists of eight items, including cough, expectoration, dyspnea, chest tightness, confidence, limitation of daily activities, quality of sleep, and levels of energy with scores ranging from 0 to 5 (0 = no impairment, 5 = greatest impairment), with higher score indicating more severe symptoms.²⁶

Calculation of the DAR and NLR

DAR= D-dimer level (mg/L)/albumin level (g/dL).

NLR= Neutrophil level/lymphocyte level.

Sample Size

The sample size was calculated by using PASS 11.0 version, yielding an optimal size of 310, considering 5% of confidence interval width (two-sided), a 95% confidence level, and odds ratio (OR)=2.8 for numbers of one-year hospital readmission ≥2. Considering a proportional distribution of 7/3, a shedding rate of 10%, a total sample size of 500 was needed for the study. Finally, the sample was acquired 567.

Statistical Analysis

Continuous data were presented as mean and standard deviation (SD) for normally distributed data and median (25%–75% quarters) for abnormally distributed data. The comparison between two groups were detected using Student’s t tests, Satterthwaite t tests, and Wilcoxon rank sum tests. Categorical data were presented as a frequency (percentage), with compared according the Chi-square tests between two groups. Potential covariates were selected according to the univariate logistic regression model and previous researches. Univariate and multivariate logistic regression models were conducted to explore the relationship between DAR, NLR and hospital readmission within one year in patients with AECOPD. The relationship was further investigated in different mMRC, CAT, COPD course, pneumonia, glucocorticoid, and antibiotic subgroups. All results were presented with ORs and 95% confidence intervals (CIs). Receiver operating characteristic (ROC) analysis was performed to determine predictive power of the DAR for the hospital readmission within one year in patients with AECOPD. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) are 2 indexes that assess improvement in model performance accomplished by adding new markers. We added mMRC, pulmonary function, and CAT to compare the predictive capability with DRA on AECOPD hospital readmission (Table S1-3 in supplementary material). All analyses were performed based on the R version 4.2.3 (2023–03-15 ucrt), and statistical significance was set at P<0.05.

Results

In total, 509 patients were included for further analysis. Patients were excluded of those without diagnosis of AECOPD (n=8), missing data on D-dimer (n=15), albumin (n=17), hsCRP (n=16), and blood urea nitrogen (n=2). The patient selection process is shown in Figure 1. With a mean age of 65.07 (10.32) years, and the majority of patients (84.68%) were males. The main clinical events for hospital readmission occurred in 117 (22.99%) patients with AECOPD. Statistical significances were found between the two groups on DAR, NLR, D-dimer, albumin, drug used, heart failure, RDW-CV, FEV1/FVC, treat with glucocorticoid, treat with antibiotic, inpatient time, PaCO₂, comorbidity, times of acute exacerbation in previous one year, mMRC, CAT, and COPD course. Table 1 shows the characteristics between the two groups.

Table 1 Characteristics of Patients with AECOPD

Figure 1 Selection process of included AECOPD patients.

Figure 2 illustrates the predictive power of DAR on the hospital readmission in patients with AECOPD. The area under the curve (AUC) value of the DAR was found to be 0.726, and 71.8% sensitivity, 66.3% specificity, 38.9% positive predictive value and 88.7% negative predictive value were reached with a DAR cut-off value of 2.21. For NLR, the cut-off value was 3.69.

Figure 2 The ROC plot of DAR and AECOPD hospital readmission.

Then, the relationship between DAR, NLR, and AECOPD hospital readmission was explored (Table 2), stratified based on the optimal cut-off value of DAR and NLR. After adjusting age, sex, mMRC, FEV1/FVC, and COPD course, DAR ≥2.21 was associated with higher odds of AECOPD hospital readmission (OR=1.80, 95% CI: 1.05–3.17). No relationship was found between NLR and odds of AECOPD hospital readmission (OR=1.57, 95% CI: 0.99–2.50).

Table 2 The Associations Between DAR, NLR and AECOPD Hospital Readmission

Table 3 shows the associations of DAR, NLR, with AECOPD hospital readmission were further explored in different mMRC, CAT, COPD course, and pneumonia populations. After adjusting covariates, DAR ≥2.21 was related to increased odds of AECOPD hospital readmission in patients of those mMRC ≥2 (OR=2.15, 95% CI: 1.18–4.10), CAT >20 (OR=2.70, 95% CI: 1.38–5.66), COPD course <10 years (OR=2.44, 95% CI: 1.14–5.71), and pneumonia (OR=2.44, 95% CI: 1.29–4.92). NLR ≥3.69 was associated with higher odds of AECOPD hospital readmission in patients of those mMRC ≥2 (OR=1.75, 95% CI: 1.07–2.90) and COPD course ≥10 years (OR=2.72, 95% CI: 1.38–5.57).

Table 3 The Associations Between DAR, NLR and AECOPD Hospital Readmission in Subgroups

Discussion

Our study suggested that high DAR was associated with increased odds of AECOPD hospital readmission within one year. Moreover, the findings indicated DAR may have superior predictive value for AECOPD hospital readmission compared to the NLR in patients with AECOPD. The relationships between DAR or NLR and AECOPD readmission risk were influenced by the duration of COPD. Higher DAR was related to higher odds of AECOPD readmission within one year in patients with CODP <10 years, while NLR was the opposite.

DAR is a novel indicator for patients with AECOPD, while it has been established that D-dimer and albumin alone can serve as predictive factors for prognosis in patients with AECOPD. D-dimer was an independent risk factor for in-hospital and 1-year death for AECOPD patients.²⁰ Low serum albumin level was associated with prolonged lengths of stay.²³ Kamstrup et al also reported in patients with moderate-to-severe COPD, a high level of D-dimer was more likely to die.²⁸ NLR could be as biomarkers in patients with AECOPD.²⁹ Our findings indicated that DAR may be a more robust predictor of AECOPD hospital readmission within one year compared to NLR, underscoring the importance of considering multiple biomarkers in assessing disease prognosis. Furthermore, future studies are needed to detect the predictive value of DAR combined with other biomarkers.

The mechanism underlying the relationship between DAR and AECOPD hospital readmission is likely multifactorial. Elevated DAR levels reflect a state of systemic inflammation and fibrinolysis, which are integral processes in the pathogenesis of COPD exacerbations.¹³ Heightened fibrinolytic activity, as indicated by elevated D-dimer levels, may exacerbate airway obstruction and impair gas exchange, thereby exacerbating respiratory symptoms and necessitating hospitalization.³⁰ Additionally, impaired albumin synthesis in patients with COPD may contribute to increased DAR levels, as albumin plays a crucial role in regulating inflammation and maintaining vascular integrity.³¹ Finally, the association between DAR and AECODP hospital readmission may be attributed to the systemic consequence s of COPD exacerbations. Acute exacerbations elicit a cascade of pro-inflammatory cytokines and acute phase reactants, including fibrinogen and D-dimer, which perpetuate systemic inflammation and endothelial dysfunction.³² This systemic inflammation not only exacerbates respiratory symptoms but also contributes to multi-organ dysfunction, increasing the risk of subsequent exacerbations and hospital readmission.

The differing associations between DAR and NLR with AECOPD hospital readmission according to COPD duration suggest potential variations in disease pathophysiology over time. In patients with shorter COPD durations, elevated DAR may be as a more sensitive marker of disease severity and exacerbation risk, reflecting early inflammatory and coagulation and abnormalities. Conversely, in individuals with long-standing COPD, NLR may become a more prominent predictor of exacerbation outcomes, reflecting chronic inflammation and immune dysregulation.^33,34 These divergent associations underscore the dynamic nature of considering disease duration in biomarker interpretation.

The findings underscore the potential of the DAR as a valuable adjunctive tool in assessing readmission risk for patients experiencing AECOPD. While DAR demonstrated an association with increased odds of readmission, it is essential to contextualize these results within the broader scope of COPD management. Clinicians should be aware that while high DAR levels could signify a higher risk of subsequent hospitalization, this biomarker does not replace clinical judgment or established management protocols. Therefore, unnecessary routine laboratory testing solely for DAR measurement is not recommended. Instead, our findings may inform clinicians in refining risk stratification and optimizing follow-up care for high-risk patients, ultimately enhancing patient outcomes through tailored post-discharge strategies.

Several limitations should be acknowledged for our study. First, this is a retrospective and single-center study which limited the number of cases. Prospective studies and random control trails are required for further evaluation of our findings. Second, although some covariates were adjusted, other possible confounders such as changes in treatment regimens after discharge were not available. Finally, no data on respiratory rehabilitation, nutritional therapy, and differences in economic or educational level, which limited the exploration of their relationship with readmission.

Conclusion

Higher DAR was associated with elevated odds of AECOPD hospital admission within one year. DAR may be a helpful biomarker for identifying AECOPD patients at higher risk of hospital readmission within one-year. Future investigations should be untaken to apply this prediction tool in order to improve patient care to reduce the need for AECOPD readmissions.